Development of transitivity in a language without/with ...
Transcript of Development of transitivity in a language without/with ...
Selected Papers of the 21st International Symposium on Theoretical and Applied Linguistics (ISTAL 21), 186-210
2016, ISSN 2529-1114, © N. Lavidas
Development of transitivity in a language
without/with object clitics:
English vs. Greek (A diachronic contrastive study)
Nikolaos Lavidas
Aristotle University of Thessaloniki
Abstract
In this historical study, I test the hypothesis of a relationship between the changes in the
null definite objects and the development of the D-system and clitics. In Lavidas (2013), I
have shown that the grammaticalization of the viewpoint aspect in English has progressed
in parallel with activity/event-noun cognate objects, but it does not appear to have
affected the development and loss of null object constructions in English. The diachronic
examination of null objects in the present study searches for the absence/presence of a
link between the emergence of new types of personal pronouns in the history of a
language without clitics, English, and in the history of a language with clitics, Greek. The
paper offers new data from the historical corpora of English and Greek, and analyzes both
the grammaticalization of the D-system in English and Greek and its link to the loss of
null definite objects in both languages. The loss of the definite null objects in English
follows the changes in the D-system but does not directly follow the development of the
personal pronouns. The relationship between the development of personal pronouns in
the object position, the changes in the D-system, and the loss of definite null objects is
evident for Greek.
Keywords: null objects, D-system, language change, English, Greek
1. Introduction
Cross-linguistically, transitive verbs can appear with an empty/null object, whereas
intransitive verbs can accept objects (see examples in 1a and 1b–d). Hence, it appears that
transitivity cannot be described as a lexical characteristic; transitives can omit their object
and intransitives can accept an object. For instance, unergatives can accept cognate
objects, measure DPs, or even regular DPs (Massam 1990). According to the traditional
approach, a transitive verb subcategorizes for and selects a direct object. The traditional
approach to objects is different from the approach followed for subjects where the subject
doi: https://doi.org/10.26262/istal.v21i0.5225
Development of transitivity in a Language without/with object clitics 187
position (EPP feature) is always present in the structure and dissociated from the
selection of the subject and the interpretation of its role (Pirvulescu & Roberge 2005). An
optimum solution to this problem would be to assume a single transitive lexical entry for
all verbs and to derive intransitive uses through a theoretical approach to null objects
(Cummins & Roberge 2004, 2008). In this way, transitivity is a (universal) syntactic
characteristic (not an idiosyncratic characteristic) (Cummins & Roberge 2004, 2008;
Lambrecht & Lemoine 1996; Larjavaara 2000). According to this idea, and similarly to
Borer (2004),1 the projection of arguments does not depend on the characteristics of the
lexical items.
(1) a. There are those who annihilate Ø...with violence - who devour Ø. (Cummins &
Roberge 2004)
b. She slept a restless sleep.
c. She ran a good race.
d. She ran a mile.
If transitivity is a (universal) syntactic characteristic, null objects are significant for
any theoretical approach. Null objects can have (i) a generic interpretation or (ii) a
definite/specific anaphoric function (see examples in 2). Cummins & Roberge (2004)
have related the relevant examples of null objects to a distinction between two types of
null objects: pro (following Sportiche 1992, 1998, for clitics; see also examples in [5])
and null N type.
(2) i. During my sabbatical I mainly intend to read Ø.
ii. A: Do you want this book?
B:*Oh! I’ve already read Ø. (Cummins & Roberge 2004, 2008)
Example (2ii B) is ungrammatical in Modern English
grammatical in Old English
grammatical in Modern French: a language with object clitics
ungrammatical in Modern Greek: although Modern Greek is
again a language with object clitics
grammatical in Ancient Greek
Table 1. Availability of definite null objects cross-linguistically and diachonically
1 See also Hale & Keyser (2002); according to them, denominal unergatives are transitive VPs.
188 Nikolaos Lavidas
Modern English and Modern Greek allow the null N type of objects. On the other
hand, examples in (3b) and (4b) need reference to an antecedent; they need the use of a
pronoun in Modern English and Modern Greek (see 5a and 5b). Old English2, Ancient
Greek3, as well as Modern French (and Modern Slavic languages, Portuguese, Chinese,
among other languages; see below) allow the null pro type of null objects, too (Pérez-
Leroux, Pirvulescu & Roberge 2006, 2008, 2011; Authier 1992).
(3) Modern English
a. -We have to get rid of all the ugly dishes before your date arrives.
-Okay, you wash N and I’ll dry N. (Goldberg 2001: 515)
b. -What did you do with the dirty dishesi?
-*She washed proi and I dried proi.
(4) Modern Greek
a. -Prepi na ksefortothume ta aplita piata prin erthi i fili su.
-Egine. Esi plenis N ke egho skupizo N.
b. -Ti ekanes ta aplitai?
-*(Ta) eplina proi ke *(ta) proi skupisa.
(5) a. She washed them and I dried them.
b. Ta katharise, ta eplina.
Modern English and Modern Greek -in contrast to Old English and Ancient Greek
(and Modern French; see section 2 for a more analytical presentation of characteristics
but also differences between languages that allow null objects) do not show an empty
pronoun category. From the above discussion, one can conclude that the default universal
transitivity property includes the availability of (the type of) null bare N as an object: V
systematically merges with an object, but this object can be a bare N and lack lexical
features (only being a semantic hyponym of V). Table 2 presents Cummins & Roberge‘s
(2004) analysis of the two basic types of null objects. A rich number of studies have
shown that children demonstrate a stage of acquisition when objects are optional (and
many null objects appear in the children data) (Jakubowicz et al. 1997; Wexler, Gavarró
& Torrens 2004, among others). It seems that children start with a broad syntax, which
2 See, for instance, Visser (1963: 525) for object drop in Old English.
3 See Luraghi 2003, 2004.
Development of transitivity in a Language without/with object clitics 189
can contain all types of null objects, and develop to a grammar that can only have generic
null Ns in the case of languages such as Modern English or Modern Greek.
Reference referential non-referential
Antecedent reference to antecedent no reference to antecedent
Contextual
recovery
recovered
by clitic
clitic-drop
("delinked")
deictic contextual
clues
present
contextual
clues
absent
Syntax pro N
Semantics features
of clitic
via
antecedent
via deixis lexical semantics of V
(i.e. Null Cognate
Object)
Pragmatics I-principle
on
linguistic
context
I-principle on
extralinguistic
context
I-principle
(less
stereotype,
more
context)
I-principle
(more
stereotype,
less
context)
Table 2. Types and characteristics of null objects (Cummins & Roberge 2004)
The aim of this historical study is to test the hypothesis of a relationship between null
definite objects and D-system and clitics (see above examples that are recovered with the
use of a clitic: She washed them and I dried them; Ta katharise, ta eplina). In Lavidas
(2013), I have shown that the grammaticalization of the viewpoint aspect in English has
progressed in parallel with the activity/event-noun cognate objects, but it does not appear
to have affected the development and loss of null object constructions in English.4 The
diachronic examination of null objects in the present study will search for the absence/
presence of a link between the emergence of a new type of personal pronouns in the
history of a language without clitics (and actually without clitics in any of its stages),
4 Due to space restriction, I will have to avoid any detail on other hypotheses about the development of null
objects; I refer to Lavidas (2013) where other possible hypotheses are examined and the relation between
the development of aspect and the development of null objects is not confirmed in contrast to the relation
between the development of aspect and the development of cognate objects.
190 Nikolaos Lavidas
English, and in a language with clitics from an early point in its history, Greek. The paper
offers new data on the comparison between the development of pronouns (= D elements;
see below), articles (= D elements too; see below) from historical corpora of English and
Greek, and further remarks on their comparative analysis. Section 2 discusses the
hypothesis of the present study—the connection between the development of null objects
and the changes in the D-system. In Section 3, the results of a corpus study are presented
and discussed. Section 3.1 shows that the grammaticalization of the D-system in English
is a later phenomenon and that the loss of null objects follows the development of the D-
system, but not the changes in the frequency of the use of personal pronouns (this fact can
be related to the absence of clitics in English). On the other hand, the relation between the
development of the personal pronouns in the object position (and the clitics) and the
changes in the D-system is evident for Greek (Section 3.2). Section 4 summarizes the
main conclusion of this study.
2. D-system and null objects: A hypothesis for the diachrony of null objects
The hypothesis of this study is that, contrary to cognate objects and their development
(Lavidas 2013), the absence of null objects is a phenomenon related to their formal status
as D-elements (Tsimpli 1999, 2003, 2004; Kowaluk 2001). Hence, the development of
the English and Greek D-system should reflect the development of null objects. This
argument is based on cross-linguistic observations: Slavic languages or Turkish, for
instance, do not have definite determiners (Kornfilt 1997; Franks & Holloway King
2000) and allow null objects (Turkish, Russian, and Polish allow null objects with a
specific reference) (Kowaluk 2001).5 For Kowaluk, the object drop in Slavic languages is
allowed because of the absence of a D-system that poses spell-out constraints on formal
features (see Franks & Holloway King [2000], for a similar suggestion). As shown in
Kowaluk, languages without object agreement (such as Slavic languages—and we can
add Old English or Ancient Greek, too) need a different approach to analysis than the one
used for Chinese or Japanese (according to Huang [1984]: ―discourse-oriented‖
languages, such as Chinese or Japanese, allow object drop). Furthermore, in Slavic
5 Cf. Vincent (1997): the common source for determiners and pronominals in Proto-Indo-European speaks
for a categorical relation between determiners and pronouns.
Development of transitivity in a Language without/with object clitics 191
languages, in contrast to discourse-oriented languages, object drop is available within
island constructions, too (and this also holds true for Old English again).6
Modern Polish,
Modern Russian, and Modern Turkish allow null objects, do not show object agreement,
and lack articles; that is, they lack a grammaticalization of the [+/-Def] features on D.
This, of course, is related to the status of pronouns in these languages, too. According to
Kowaluk (following Tsimpli & Stavrakaki [1999]), it is logical to assume that languages
that lack [D] should have pronouns represented not as (functional) DPs, but as (lexical)
NPs.
Hence, the distinction and the hypothesis for a diachronic development, according to
Kowaluk‘s (and Tsimpli & Stavrakaki‘s) approach would be between languages that:
(a) have clitics that are D elements and strong pronouns that are N elements; these
languages do not allow null definite/referential objects; and
(b) do not have D elements (lack articles), but may have full personal pronouns
and clitics of N-type (according to Kowaluk‘s analysis for Modern Polish); these
languages allow null definite/referential pronouns.7
Note that pre-Modern Greek clitics are similar to Modern Polish clitics in a clear
aspect—the optionality of positions where they appear (see the discussion below of the
development of clitics in Greek). Hence, one can hypothesize that neither Modern Polish
6 See examples in Denison (1993: 189):
(1) a. het hiene tha niman & thaeron bescufan
ordered him then take and therein cast
‗He ordered him then to be taken and case inside‘ (Or. 34.13)
b. ic Beda Cristes theaw... sende gretan thone
I Bede Christ‘s servant send greet the
leofastan cyning & halettan Ceowulf
most-beloved king and honor Ceowulf
‗I, Bede, Christ‘s servant, send people to greet and honor the most beloved King Ceowulf‘.
(Bede 417.7) 7 Modern Polish NPs for Kowaluk are as follows in (1). A hypothesis can be that NPs in Old English and
Ancient Greek are also of the same type. See below for a discussion of Old English and Homeric and
Classical Greek NPs. Modern Polish NPs have deictic and referential characteristics; they do not
distinguish between definite and indefinite but between animate and inanimate.
(1)
192 Nikolaos Lavidas
nor pre-Modern Greek pronouns are ―severely deficient clitics‖ as they are in Modern
Standard Greek. We follow Kowaluk and claim that these (clitic) pronominals of Modern
Polish (and pre-Modern Greek) are N-type elements that do not need to attach to a
(verbal) functional head. The empty objects in these cases (Modern Polish and pre-
Modern Greek) are empty pronominals too.
Modern Polish No D-system8 N-type pronouns empty pronominal objects
Modern Greek D-system severely deficient clitics *empty pronominal objects
(grammatical clitics, attached
to the verbal functional head)
Modern English D-system No clitics *empty pronominal objects9
Table 3. D-system, clitics, and null objects cross-linguistically and diachronically:
Distribution and hypotheses
Following the above remarks, our hypothesis is that null objects in Old English and
Ancient Greek and their loss could be accounted for in a similar way, in terms of
emergence of a new D-system. In contrast to Slavic languages or Turkish, Modern Greek
has a D-system with definite and indefinite determiners (their inflection expresses case
and phi-features).10
Tsimpli (2003) has shown that Modern Greek, for instance, has clitics
and determiners that share all other formal features, except for person.
8 ―No D-system‖ or ―D-system‖ here means absence or presence of a D-system that poses spell-out
constraints on formal features. 9 And the hypothesis would add the following:
Pre-Modern Greek No D-system N-type pronouns empty pronominal objects
Pre-Modern English No D-system N-type pronouns empty pronominal objects 10
Note that D-system and pronoun are interrelated: in either way of analysis of the pronouns (analysis of
movement or base-generated pronouns), the pronouns are in D:
(i)
(ii)
Furthermore, there can be no co-existence of articles and pronouns:
(1) *o ego; *the I (2) *i esi; *the you
Development of transitivity in a Language without/with object clitics 193
Article Clitic
Subject Object Subject Object
Singular o (masc.), i
(femin.), to
(neuter)
ton, tin, to Ø ton, tin, to
Plural i, i, ta tus, tis, ta Ø tus, tis, ta
Table 4. Articles and clitics in Modern Greek, from Tsimpli (2003)
In Modern English, the lack of marking of case and phi-features has probably blocked
the emergence of clitics that would be associated with definite articles.
Modern English: definite determiners with no marking of case, gender, or number
(definite article: the)
no clitics
Modern Greek: definite determiners with marking of case, gender, number
clitics
Table 5. Articles and clitics: Modern English vs. Modern Greek
According to Tsimpli & Stavrakaki (1999), definite articles and object clitics in
Modern Greek are both associated to referentiality that derives from a feature-matching
relation between the Def(initeness) head (for the Modern Greek articles) and an
antecedent (for the Modern Greek object clitics):
(6) [DefP... [DP[case]/[phi-]...[NP...]]] (from Tsimpli 2003)
On the other hand, indefinite articles and demonstratives in Modern Greek are
inherently specified with the feature of referentiality, and they appear on Def. head
(Tsimpli 2003). Modern Greek indefinite articles (and demonstratives) consist of
interpretable features. Furthermore, there is a contrast between personal pronouns of 1st
and 2nd
person that have phi-features (interpretable person feature) and 3rd
person
194 Nikolaos Lavidas
personal pronouns that have uninterpretable features (and can vary cross-linguistically
and diachronically) (Cardinaletti & Starke 1999; Tsimpli & Stavrakaki 1999; van
Gelderen 2011, among others).
Accordingly, in our corpus study below, we will distinguish between definite and
indefinite articles as well as between 3rd
and 1st/2
nd person object pronouns (and clitics for
Greek). Note that with regard to language change, interpretable features tend to change
into uninterpretable—except for the case of replacement of an element (where we can
have the emergence of a new element with interpretable features) (see van Gelderen
2000, 2011). For English, the changes in the D-system have been analyzed as a reanalysis
of interpretable features into uninterpretable in the case of determiners (change of
demonstratives into determiners). With regard to pronouns, the Old English pronouns are
replaced by a new system of pronouns (through contact with Scandinavian languages)
(van Gelderen 2000, 2011). (See below for details).
To summarize, pronouns, and mainly clitics, and definite articles are associated with
uninterpretable formal features (case and phi-features). With regard to object clitics,
modern languages with null objects that lack a D-system (that poses spell-out constraints
on formal features), such as Slavic languages or Turkish, lack pronominal clitics, too.
This can hold for Old English (see the discussion and the relevant bibliography below); it
may hold for Ancient Greek (see the discussion below), but it also shows that a ―new‖ D-
system, of course, does not mean the same new D-system for English and Greek. English
developed a new D-system without clitics and Greek a new D-system with clitics. Both
languages lost the null definite objects after the emergence of the new D-system.
3. A corpus study: Development of articles and personal pronouns (in the object
position) in English and Greek
I conducted a corpus search of all texts in the York-Toronto-Helsinki Parsed Corpus of
Old English Prose (YCOE), the Penn-Helsinki Parsed Corpus of Middle English
(PPCME2), and the Penn-Helsinki Parsed Corpus of Early Modern English (PPCEME)
(Kroch & Taylor 2000; Taylor et al. 2003; Kroch, Santorini & Delfs 2004; Kroch,
Santorini & Diertani 2010) to test the major hypothesis presented above in this paper.
Furthermore, to compare the results from the English data, a language without clitics, to
Development of transitivity in a Language without/with object clitics 195
the Greek data, a language with clitics, I conducted a corpus search in the PROIEL
annotated corpus and in the material found in the Perseus under Philologic (Homer,
Herodotus, New Testament).11
The purpose of the corpus searches was to count:
(i) personal pronouns (and clitics for Greek) in object position;
(ii) the contrast between 3rd
vs. 1st and 2
nd person personal pronouns;
(iii) definite vs. indefinite articles were examined since the hypothesis will be that
change in D-system affects the presence of null objects; and
(iv) full NPs/DPs in object position.
3.1 The English data
In Lavidas (2013), I have presented the stages of development of null objects. Since there
is no coding for null objects (in contrast to null subjects) in the corpora, I have decided to
count the presence/absence of overt direct objects in the relevant corpora. A Pearson chi-
square test was performed to assess the relationship between the periods and the
development of overt direct objects. The results were statistically significant for the
comparison between O4 and M1 (χ2=35.46, p<.001), with an effect size of φ=.42, which
is a large size effect.12
The corpora (see Figure 1) confirm Visser‘s (1963) remarks
regarding the increase in the presence of overt direct objects during the Middle and Early
Modern English periods, but the primary change in overt direct objects can be observed at
the beginning of the Middle English period (Lavidas 2013).
11
For English, the search was carried out automatically using the program Corpus Search 2 (Randall 2005-
2007). 12
The chronological periods (according to the York, Helsinki, and Pennsylvania corpora) in all figures and
tables for English are the following: O2 (Old English; 850–950); O3 (Old English; 950–1050); O4 (Old
English; 1050–1150); M1 (Middle English; 1150–1250); M2 (Middle English; 1250–1350); M3 (Middle
English; 1350–1420); M4 (Middle English; 1420–1500); E1 (Early Modern English; 1500–1569); E2
(Early Modern English; 1570–1639); E3 (Early Modern English; 1640–1710); ); MBE (Modern British
English; 1700-1914).
196 Nikolaos Lavidas
Figure 1. Clauses with an overt direct object vs. clauses without an overt direct object in
the history of English (data from Lavidas 2013)
With regard to the D-system and its development in English, Figure 2 presents the
emergence of definite articles in the first period of Middle English. Indefinite articles
have a ―modern‖ frequency only after 1350, as seen in Figure 3. A Pearson chi-square
test was performed to assess the relationship between the different periods of the history
of the English language and the development of the definite and indefinite article. The
results with respect to the use of the definite article were statistically significant for the
comparison between O4 and M1 (χ2=19.672, p<.001), with an effect size of φ=.314,
which is a medium size effect.13
Figure 2. Use vs. absence of definite articles in the history of English
13
Due to space restrictions, the discussion of the statistical results in the whole paper has to be indicative
and short.
Development of transitivity in a Language without/with object clitics 197
Figure 3. Use of definite articles and use of indefinite articles (compared to the total
number of nouns) in the history of English
Figure 4 shows a variation in the frequency of the use of 1st and 2
nd person personal
pronouns that appears to follow the tendencies for the overt objects (see above). The 3rd
person personal pronouns, however, appear to be stable in their frequency in the different
periods of the history of the English language. In Figure 5, the main observations with
regard to the development of articles and 3rd
person personal pronouns are presented in a
comparative way: the frequency of the 3rd
person personal pronouns in the accusative
case is stable, whereas the emergence of the definite articles is shown for Early Middle
English. A Pearson chi-square test was performed to assess the relationship between the
different periods of the history of the English language and the development of the
personal pronouns in the object position. The results with respect to the 1st/2
nd person
personal pronouns in object position were statistically significant for the comparison
between O4 and M1 (χ2=18.233, p<.001), with an effect size of φ=.302, which is a
medium size effect.
198 Nikolaos Lavidas
Figure 4. Frequency of use of 3rd
person personal pronouns in the accusative/objective
vs. frequency of use of 1st and 2
nd person personal pronouns (compared to the total
number of NPs in the accusative/objective) in the history of English
Figure 5. Frequency of use of definite articles and frequency of 3rd
person personal
pronouns in the accusative/objective in the history of English
Changes in the D-system of English, obviously represented in the results of the corpus
study, include the following: in late Old English, the demonstrative pronouns (see Table
6) are reanalyzed as definite articles (van Gelderen 2000, 2011); and in late Old English
(middle of the 12th
century), a new 3rd
person plural personal pronoun (they) and a new
3rd
person feminine singular (she) are attested (van Gelderen 2000, 2008, 2011).
Development of transitivity in a Language without/with object clitics 199
Singular,
Masculine
Singular,
Feminine
Singular,
Neuter
Plural
Nom. se seo þæt þa
Gen. þæs þære þæs þara
Dat. þæm þære þæm þæm
Acc. þone þa þæt þa
Table 6. Demonstratives in Old English14
For van Gelderen (see also the results from the corpus search above), the English D-
system changed at the end of the Old English period. This can be clearly shown in the
changes from the Preface of the Anglo-Saxon Chronicle that included no articles and few
demonstratives, to a part from the year 1130 of the Chronicle with articles and
demonstratives (and an increase of their use), to a part from the year 1137 with clear
articles and new personal pronouns (replacing the old system of personal pronouns).
According to van Gelderen, Old English personal pronouns are not deictic/referential.
Early Middle English pronouns are deictic/referential. Old English personal pronouns are
also attested with a reflexive interpretation. Pronouns of Old English could be described
as having characteristics of clitics, according to Pintzuk (1996), but this probably means
their reduced referentiality because they do not show clitic behavior (Bech 2001, among
others). The new 3rd
person plural and the new 3rd
person feminine singular pronouns are
attested in the middle of the 12th
century.
Singular Plural
1st Nom. ic we
Gen. min ure
Dat. me us
Acc. mec usic
14
se (masculine nominative) is reanalyzed as the definite article the in later English, and þæt (neuter
nominative) is reanalyzed as the demonstrative that (Wood 2007). The demonstrative pronouns in Old
English are not definite articles because the demonstrative pronouns are in complementary distribution with
the possessive pronouns (and until the end of the Old English period).
200 Nikolaos Lavidas
2nd
Nom. þu ge
Gen. þin eower
Dat. þe eow
Acc. þec eowic
3rd
Nom. he/heo/hit hi
Gen. his/hire/his hira
Dat. him/hire/him him
Acc. hine/heo/hit hi
Table 7. Personal pronouns (singular and plural) in Old English
Table 8 summarizes the changes in articles and pronouns in the history of English. For
these changes in terms of features, see Figure 6 from van Gelderen (2011) (van Gelderen
provides a detailed presentation of the changes).
Old English: ―articles‖ = demonstratives = deictic/referential
Early Middle English (middle of the 12th
cent.): definite articles and demonstratives
Old English: ―personal pronouns‖ = demonstratives
Early Middle English: personal pronouns = deictic/referential
Table 8. A summary of the changes in English
Demonstrative
[i-phi] = person and deixis
[i-loc]
article pronoun C (that Complementizer)
[u-phi] [i-phi]
Figure 6. Feature changes in English; van Gelderen (2011)
Development of transitivity in a Language without/with object clitics 201
If the D-system is responsible for the change as our hypothesis states, then the
different status of articles and pronouns should have affected the subject (its presence and
omission), too. Indeed, Old English allows subject pro-drop as well as the use of personal
pronouns to express the continuing topic and the use of personal pronouns with a
reflexive interpretation (see, for instance, Walkden 2013).
If null objects are caused by the changes in the D-system, then object (and subject)
drop would follow the development of D, both in English and Greek. The corpus study
reveals very interesting results: it appears that there is a corresponding increase in the use
of personal pronouns in the object position (and mainly of the 1st/2
nd person personal
pronouns), but not an increase of the 3rd
person personal pronouns for English, as
expected (if we distinguish them from the 1st/2
nd person pronouns).
3.2 The Greek data
In this section, I will present the corpus study that I conducted for Greek. There are no
previous studies on the development of null objects in the history of the Greek language.
The problem is again that the coding for null objects is not a simple task, and it is not
available in large corpora of Greek. For this reason, I counted overt direct objects in the
accusative case in Homeric Greek (Odyssey and Iliad), Herodotus, and the Greek New
Testament.15
Note that the tendency of loss of the null objects is evident even though
other parameters obviously interfere in this case, such as the increase of the number of
accusatives in later Greek (Figure 7).
15
I conducted a corpus search in the PROIEL annotated corpus and in the material found in the Perseus
under Philologic (Homer, Herodotus, New Testament). I also made use of the ANNIS corpus search tool.
202 Nikolaos Lavidas
Figure 7. Development of presence of overt direct objects in Greek
Figure 8 presents the emergence of the definite article in Classical Greek, and the
contrast between Classical and Homeric Greek with regard to articles; there is no regular
use of the definite article in Homeric Greek. The situation in New Testament Greek is
very similar to Classical Greek. In Figure 9, the emergence of indefinite articles is also
represented; it appears that the emergence of indefinite articles is completed in a later
stage in the Greek New Testament. A Pearson chi-square test was performed to assess the
relationship between the different periods of the history of the Greek language and the
development of definite and indefinite articles. The results with respect to the new
definite articles were statistically significant for the comparison between Homer and
Herodotus (χ2=50.439, p<.001), with an effect size of φ=.502, which is a large size effect.
Figure 8. Use of definite articles vs. absence of definite articles in Greek
(Homer, Herodotus, New Testament).
Development of transitivity in a Language without/with object clitics 203
Figure 9. Use of indefinite articles (compared to all NPs) in Greek (Homer, Herodotus,
New Testament)
Figure 10 presents a different development for 3rd
person personal pronouns in Greek
than in English: probably because of the clitic behavior (from a certain period and then)
of personal pronouns in Greek in contrast to English. Hence, the use of 3rd
person
personal pronouns in Greek is not stable and there appears to be an increase in their use in
Classical Greek, and mainly in the Greek New Testament. 1st and 2
nd personal pronouns
follow the increase of 3rd
person pronouns, but with a small delay (the Greek New
Testament is the text where a high increase in the frequency of the use of 1st and 2
nd
person personal pronouns can be observed). In Figure 11, similar tendencies for definite
articles and personal pronouns in Greek (but not in English—see above) are evident: the
increase in the frequency of the definite articles is attested in a parallel path with the
increase in the use of the 3rd
person personal pronouns (and the change of the personal
pronouns into clitics) in Greek. A Pearson chi-square test was performed to assess the
relationship between the different periods of the history of the Greek language and the
development of the use of personal pronouns in the object position. The results with
respect to the 3rd
person personal pronouns were statistically significant for the
comparison between Homer and Herodotus (χ2=5.018, p<.05), with an effect size of
φ=.158, which is a small size effect. The results with respect to the 1st/2
nd person personal
pronouns were statistically significant for the comparison between Herodotus and the
204 Nikolaos Lavidas
New Testament (χ2=6.575, p=.01), with an effect size of φ=.181, which is a small size
effect.
Figure 10. Frequency of use of 3rd person personal pronouns in the accusative vs.
frequency of use of 1st and 2nd person personal pronouns in the accusative (compared to
the total number of NPs-accusative): Homer, Herodotus, and the New Testament
Figure 11. Use of definite articles and 3rd
person personal pronouns in the accusative in
Homer, Herodotus, and the New Testament
The definite article is assumed (in the relevant bibliography) to have developed
between Homer and Classical Greek, whereas the indefinite article appears in Hellenistic
Greek (first cent. AD) (Guardiano 2003, 2013; Manolessou 2001; Bakker 2009). The
etymology of the Greek definite article connects it to an Indo-European demonstrative
(see the relevant remarks for English, above: a change is evidenced from demonstrative
into definite article, for English, too). In Homer, no definite (or indefinite) article is
Development of transitivity in a Language without/with object clitics 205
present. The anaphoric/non-anaphoric, count/mass, kind/existential, and proper/common
nouns are attested as bare Nouns.
(7) éntha kaì ēmatíē mèn huphaínesken mégan histón
‗Then day by day she would weave at the great web.‘ (Hom. Od. 2, 104; from
Guardiano 2013)
Guardiano (2013: 78) observed that Homeric Greek is similar to many other Indo-
European languages and that ―for instance [...] the spoken Slavic varieties observed by
Trovesi (2004) [...] make systematic use of a phonologically reduced form of a distal
demonstrative in typically definite contexts [...].‖ Note that, according to our basic
hypothesis, Slavic languages allow null objects because they do not have a D-system with
definite/indefinite determiners. Homeric Greek ho/hē/tó appear in some definite contexts,
but not systematically, and do not occur with a demonstrative, showing characteristics of
demonstrative pronouns themselves (Guardiano 2013). Hence, definiteness is not
grammaticalized in Homeric Greek, and all (types of) Ns can be bare. Classical Greek
and New Testament Greek ho/hē/tó, on the other hand, are systematically present in all
definite contexts (and bare nouns are not definite; see Guardiano 2011, 2013 for details).
Definiteness is grammaticalized in Classical and New Testament Greek, but there is not a
distinction between a definite determiner and an indefinite determiner with singular count
nouns until the first century A.D. After the first century A.D., a definite article occurs in
all definite contexts even with proper nouns, whereas an indefinite article becomes
available. With regard to personal pronouns in the history of Greek, most of the studies
examine aspects of the development of clitics in Greek with regard to their position (from
post- to preverbal in Modern Standard Greek). We would like to argue that the first
change in the history of Greek pronouns, the change that resulted in the ―verb-centrality‖
of pronouns, should be distinguished from the other phonological changes (as they have
been correctly described) that refer to the post- or preverbal position of the clitic. We
would like to claim, then, that in Greek (similar to English, see above) we first have a
categorical change in the pronouns from demonstratives into referential pronouns. The
subsequent changes that concern the position of the pronouns are possible because -in
contrast to English- Greek marks case (and phi-features) in object pronouns and their
206 Nikolaos Lavidas
function can be recoverable even if their position changes. Hence, we have to distinguish
between the first change in Greek pronouns and the rest of the changes: object pronouns
in Homeric Greek are not related to the verb, their position follows Wackernagel‘s law
(1892), and they tend to appear together with other ―light‖ elements in a clause-second
position. So, they are not (typical) object clitics.
Janse (1993, 2008) has shown that object pronouns in Homeric Greek must be defined
with reference to intonation units, because they are attracted to foci (new or contrastive
information). What happens in Classical Greek and mainly in post-classical Greek (New
Testament Greek) is a clear syntactic change in the features of object pronouns. What
changes is not mainly their position (depending on the context, they can still appear
before [Marshall 1987; Janse 1993, 1995-96, 2008] or after verbs), but their status. Now,
object pronouns are ―verb-centered‖ and they have a new role—they are (real) objects
that refer to an antecedent. In New Testament Greek, their position depends on
phonological parameters again (focused elements can cause preverbal position of the
pronoun) (Janse 1993, 2008), but the object pronouns now have a new featural
composition and a new syntactic behavior (what can be described as verb-centered and
deictic/referential).
(8) tís mou hḗpsato?
‗Who touched me?‘ (Mark 5.31; from Janse 1993)
What follows is that object pronouns become proclitic and appear before the verb
(during the Medieval Greek period and in Modern Standard Greek): they are marked for
case, they are ―able‖ to move before V, to easily accomplish their function of referring to
an antecedent N.
The Greek data derived from the corpus study are clearer than the English data in
showing the grammaticalization of the D-system in Greek; the reason is that Greek
continued with the emergence of pronouns with clitic behavior -in contrast to English that
does not acquire clitics. However, the restrictions in the data with regard to the null
objects, and mainly the type of null object that appears in the different stages of the
history of the Greek language, does not allow the appearance of a very obvious parallel
development for the D-system and the loss of definite null objects in Greek. Further study
Development of transitivity in a Language without/with object clitics 207
that will include later texts from the history of Greek is required. What can be evidenced
in Greek, in contrast to the English situation, is a link between the development of the
personal pronouns, mainly the 3rd
person personal pronouns, and the development of the
definite articles (and the null objects).
4. Conclusion
The hypothesis of this study was that the absence of null objects is a phenomenon related
to their formal status as D-elements (Tsimpli 1999, 2003, 2004; Kowaluk 2001); the
development of the English and Greek D-system should reflect the development of null
objects. Modern languages with null objects that lack a D-system (that poses spell-out
constraints on formal features) lack pronominal clitics, too. This can hold for Old English
and Ancient Greek, but English developed a new D-system without clitics and Greek a
new D-system with clitics. Both languages lost the null definite objects after the
emergence of the new D-system. If null objects have their cause in the changes in the D-
system, then object drop would follow the development of D, both in English and Greek.
The corpus study reveals very interesting results: it appears that, for English, there is a
corresponding increase in the use of personal pronouns in the object position (mainly of
the 1st/2nd person personal pronouns), but not an increase of the 3rd person personal
pronouns, as expected. Greek presents a different development for 3rd person personal
pronouns than English: probably because of the clitic behavior of personal pronouns in
Greek, in contrast to English. The use of 3rd person personal pronouns in Greek is not
stable and there appears to be an increase in their use in Classical Greek, and mainly in
the Greek New Testament.
References
Authier, M.1992. Iterated CPs and embedded topicalization. Linguistic Inquiry 23(2): 326-336.
Bakker, S.J. 2009. The Noun Phrase in Ancient Greek. A Functional Analysis of the Order and Articulation
of NP Constituents in Herodotus. Leiden: Brill.
Bech, K. 2001. Word Order Patterns in Old and Middle English. A Syntactic and Pragmatic Study. PhD
dissertation, University of Bergen.
Borer, H. 2004. The Grammar machine. In A. Alexiadou, E. Anagnostopoulou & M. Everaert (eds), The
Unaccusativity Puzzle. Oxford: Oxford University Press, 288-331.
208 Nikolaos Lavidas
Cardinaletti, A. & M. Starke. 1999. The typology of structural deficiency: A case study of the three classes
of pronouns. In H. van Riemsdijk (ed.), Clitics in the Languages of Europe. Berlin/New York: Mouton
de Gruyter, 145-233.
Cummins, S. & Y. Roberge. 2004. Null objects in French and English. In J. Auger, C. Clements & B.
Vance (eds), Contemporary approaches to romance linguistics: Selected Papers from the 33rd
Linguistic Symposium on Romance Languages (LSRL). Amsterdam and Philadelphia: John Benjamins,
121-138.
Cummins, S. & Y. Roberge. 2008. A modular account of null objects in French. Syntax 8(1): 44-64.
Denison, D. 1993. English historical syntax: Verbal constructions. London: Longman.
Franks, S. & T. Holloway King. 2000. Oxford Studies in Comparative Syntax: A handbook of Slavic clitics.
Oxford: Oxford University Press.
van Gelderen, E. 2000. A History of English Reflexive Pronouns. Amsterdam: John Benjamins.
van Gelderen, E. 2008. Where did late merge go? Grammaticalization as feature economy. Studia
Linguistica 62: 287-300.
van Gelderen, E. 2011. The linguistic cycle. Oxford: Oxford University Press.
Goldberg, A.E. 2001. Patient arguments of causative verbs can be omitted: The role of information
structure in argument distribution. Language Sciences 34: 503-524.
Guardiano, C. 2003. Struttura e storia del sintagma nominale nel greco antico: Ipotesi parametriche. Tesi
di Dottorato, Universita Pisa.
Guardiano, C. 2011. Parametric changes in the history of the Greek article. In D. Jonas, J. Whitman & A.
Garrett (eds), Grammatical change: Origins, nature, outcomes. Oxford, Oxford University Press, 179-
197.
Guardiano, C. 2013. The Greek definite article across time. Studies in Greek Linguistics 33: 76-91.
Hale, K. & S.J. Keyser. 2002. Prolegomenon to a theory of argument structure. Cambridge: MIT Press.
Huang, J.C.-T. 1984. On the distribution and reference of empty pronouns. Linguistic Inquiry 15: 531-574.
Jakubowicz, C., N. Müller, B. Riemer & C. Rigaut. 1997. The case of subject and object omissions in
French and German. In E.h Hughes, M. Hughes & A. Greenhill (eds), Proceedings of the 21st BUCLD.
Somerville, Mass.: Cascadilla Press, 331-342.
Janse, M. 1993. La position des pronoms personnels enclitiques en grec néo-testamentaire à la lumière des
dialectes néo-helléniques. In C. Brixhe (ed.), La koiné grecque antique I. Nancy: Presses Universitaires
de Nancy, 83-121.
Janse, M. 1995-96. Phonological aspects of clisis in ancient and modern Greek. Glotta 73: 155-167.
Janse, M. 2008. Clitic doubling from ancient to Asia Minor Greek. In D. Kallulli & L. Tasmowski (eds),
Clitic doubling in the Balkan languages. Amsterdam: Jоhn Benjаmins, 165-202.
Kornfilt, J. 1997. Turkish. NY: Routledge.
Kowaluk, A. 2001. The acquisition of determiners and pronouns in English L2. Ph.D. thesis, Cambridge,
UK.
Kroch, A., B. Santorini & L. Delfs. 2004. The Penn-Helsinki Parsed Corpus of Early Modern English
(PPCEME). Department of Linguistics, University of Pennsylvania. CD-ROM, first edition,
(http://www.ling.upenn.edu/hist-corpora).
Kroch, A., B. Santorini & A. Diertani. 2010. The Penn-Helsinki Parsed Corpus of Modern British English
(PPCMBE). Department of Linguistics, University of Pennsylvania. CD-ROM, first edition,
(http://www.ling.upenn.edu/hist-corpora).
Development of transitivity in a Language without/with object clitics 209
Kroch, A. & A. Taylor. 2000. The Penn-Helsinki Parsed Corpus of Middle English (PPCME2).
Department of Linguistics, University of Pennsylvania. CD-ROM, second edition,
(http://www.ling.upenn.edu/hist-corpora).
Lambrecht, K. & K. Lemoine. 1996. Vers une grammaire des compléments zéro en français parlé. In J.
Chuquet & M. Fryd (eds), Travaux linguistiques du CERLICO 9. Absence de marques et représentation
de l'absence. Rennes: Presses universitaires de Rennes, 279-306.
Larjavaara, M. 2000. Présence ou absence de l’objet. Limites du possible en français contemporain.
Helsinki: Academia Scientiarum Fennica.
Lavidas, N. 2013. Null and cognate objects and changes in (in)transitivity. Acta Linguistica Hungarica
60(1): 69-106.
Luraghi, S. 2003. Definite referential null objects in Ancient Greek. Indogermanische Forschungen 108:
167-194.
Luraghi, S. 2004. Null objects in Latin and Greek and the relevance of linguistic typology for language
reconstruction. Proceedings of the 15th Annual UCLA Indo-European Conference, JIES Monograph 49:
234-256.
Manolessou, I. 2001 The evolution of the demonstrative system in Greek. Journal of Greek Linguistics 2:
119-148.
Marshall, M.H.B. 1987. Verbs, nouns, and postpositives in Attic prose. Edinburgh: Scottish Academic
Press.
Massam, D. 1990. Cognate objects as thematic objects. Canadian Journal of Linguistics 35: 161-190.
Pérez-Leroux, A.T., M. Pirvulescu & Y. Roberge. 2006. Early object omission in child French and English.
In J.-P. Montreuil & C. Nishida (eds), New perspectives on romance linguistics. vol. I. John Benjamins:
Amsterdam, 213-228.
Pérez-Leroux, A.T., M. Pirvulescu & Y. Roberge. 2008. A syntactic transitivity approach to null objects in
child language. Lingua 118: 370-398.
Pérez-Leroux, A.T. M. Pirvulescu & Y. Roberge. 2011. Topicalizations and object drop in child language.
First Language 31: 280-299.
Pirvulescu, M. & Y. Roberge. 2005. Licit and illicit null objects in L1 French. In R.S. Gess & E.J. Rubin
(eds), Theoretical and experimental approaches to romance linguistics. Amsterdam: John Benjamins,
197-212.
Pintzuk, S. 1996. Cliticization in Old English. In A. Halpern & A.M. Zwicky (eds), Approaching second:
Second position clitics and related phenomena. Stanford, CA: CSLI Press, 375-409.
Sportiche, D. 1992. Clitic constructions. In J. Rooryck & L. Zaring (eds), Phrase structure and the lexicon.
Bloomington, Indiana: IULC, 213-276.
Sportiche, D. 1998. Pronominal clitic dependencies. In H. van Riemsdijk (ed.), Language typology: Clitics
in the European languages. Berlin: Mouton de Gruyter, 679-708.
Taylor, A., A. Warner, S. Pintzuk & F. Beths. 2003. The York-Toronto-Helsinki parsed Corpus of Old
English prose (YCOE). Department of Language and Linguistic Science, University of York. Available
through the Oxford Text Archive.
Trovesi, A. 2004. La genesi di articoli determinativi. Modalità di espressione della definitezza in ceco,
serbo-lusaziano e sloveno. Pavia: Franco Angeli.
Tsimpli, I.M. 1999. Determiners and clitics in Greek. In A. Kakouriotis and V. Bolla‐Mavrides (eds),
Working Papers in Theoretical and Applied Linguistics 6, Aristotle University of Thessaloniki,
189‐208.
210 Nikolaos Lavidas
Tsimpli, I.M. 2003. Interrogatives in the Greek/English interlanguage: A minimalist account. In E.
Mela‐Athanasopoulou (ed.), Selected Papers on Theoretical and Applied Linguistics. Thessaloniki:
Aristotle University of Thessaloniki, 214‐225.
Tsimpli, I.M. 2004. Features in L1 and L2 acquisition: Evidence from Greek clitics and determiners. In H.
Hendricks (ed.), Analyse comparative des processus d'acquisition en L1 et L2. Special Issue of AILE 20:
87-128.
Tsimpli, I.M. & S. Stavrakaki. 1999. The effects of a morpho-syntactic deficit in the determiner system:
The case of a Greek SLI child. Lingua 108: 31-85.
Vincent, N. 1997. The emergence of the D-System in Romance. In A. van Kemenade & N. Vincent (eds),
Parameters of morphosyntactic change. Cambridge: Cambridge University Press, 149-169.
Visser, F.Th. 1963. An historical syntax of the English language. Volume 1. Leiden: EJ Brill.
Wackernagel, J. 1892. Über ein Gesetz der indo-germanischen Wortstellung. Indogermanische
Forschungen 1: 333-436.
Walkden, G. 2013. Null subjects in Old English. Language Variation and Change 25(2): 155-178.
Wexler, K., A. Gavarró & V. Torrens. 2004. Feature checking and object clitic omission in child Catalan
and Spanish. In R. Bok-Bennema, B. Hollebrandse, B. Kampers-Manhe & P. Sleeman (eds), Romance
languages and linguistic theory 2002. Amsterdam: John Benjamins, 253-270.
Wood, J.L. 2007. Demonstratives and possessives: From Old English to present-day English. In E. Stark, E.
Leiss & A. Werner (eds), Nominal determination: Typology, context constraints, and historical
emergence. Amsterdam: John Benjamins, 339-361.